Over the past several years it has become evident that normal host cells functionally contribute to tumor initiation and progression. While there has been a wealth of studies eliciting bone marrow-derived progenitor and support cells as critical contributors to tumor angiogenesis and progression, hardly anything is known about the functional significance of stem cell populations that potentially reside within the tumor. We recently discovered a population of multipotent mesenchymal stem cell-like cells (MPMSCs) in normal pancreatic islets as well as in pancreatic islet tumors of the transgenic mouse model Rip1Tag2. This cells display expression of mesenchymal surface markers and have high self-renewal capacity as they grow in suspension as spheres - the abilities which only stem cells posses. Also MPMSC cells can differentiate in vitro into pericytes, endothelial cells and neurons. Interestingly, we found that MPMSCs recruited to tumor not from bone marrow, but probably reside within pancreatic islets and become activated during pancreas tumor progression. I hypothesized that while MPMSC cells exhibit multipotentiality in normal tissue, they might be steered to predominantly produce vascular cells (pericytes and endothelial cells) in a tumor setting to support neovascularization and repress the formation of other cell types such as neurons. The Rip1Tag2 transgenic mouse model provides a unique platform to study MPMSCs at distinct stages of disease progression as it develops pancreatic islet cell carcinomas in a multi-step pathway arising from normal beta cells in the islets of Langerhans through two distinct premalignant stages to highly vascularized large adenomas and invasive carcinomas. The mice die between 14 and 15 weeks as a combined effect of hyperinsulinemia and tumor burden. In order to test hypothesis I propose to isolate MPMSCs from different stages of pancreatic islet carcinogenesis and test whether they become activated, increase their numbers and preferentially differentiate into pericytes and endothelial cells during the course of tumor development and progression. I will also further assess these phenotypes with fluorescently-tracked MPMSCs during tumor propagation of two mice tumor models. I will co-inject MPMSC cells with 2TC3 cells (tumor cells isolated from Rip1Tag2 mice) under the kidney capsule and with melanoma (B16 cells) into dermic. Also I intended to identify the underlying signaling circuit that determines the differentiation fate of MPMSCs in normal pancreatic islets and during pancreatic tumorigenesis. While there are many signaling pathways and factors that would be worthwhile and important to assess, I decided to focus on PDGFR2 and CD271/p75 as a potential regulators of MPMSC differentiation into pericytes.